If you've ever looked at your Google Ads and Meta Ads dashboards and wondered why the same conversion shows up differently in each platform, or why your total conversions across platforms add up to more than your actual sales, you're running into attribution windows.
Attribution windows determine how much credit each ad platform gives itself for a conversion. Set them wrong — or worse, not understand what they mean — and you'll make budget decisions based on fiction.
This guide explains what attribution windows are, why the defaults are almost always wrong for your business, and how to configure them so your reporting actually reflects reality.
What Is an Attribution Window?
An attribution window is the length of time after someone clicks or views your ad during which a conversion can still be credited back to that ad.
For example, if you're using a 7-day click attribution window, and someone clicks your ad on Monday but doesn't convert until Wednesday, the conversion still counts. If they convert on the following Tuesday (9 days later), it doesn't count.
Every ad platform has attribution windows. The problem is that each platform sets different defaults, uses different logic, and claims credit differently — which is why the same customer journey gets reported as totally different conversions depending on where you look.
The Default Windows (And Why They're Broken)
Google Ads
Default: 30-day click, 1-day view
This means Google will claim credit for a conversion if it happens within 30 days of someone clicking your ad, or within 1 day of someone seeing (but not clicking) your ad.
The problem: 30 days is far too long for most businesses. If someone clicks your ad for running shoes in early April, forgets about it, then comes back via organic search in early May and buys, Google Ads will still take credit — even though the ad clearly didn't drive that purchase.
Meta Ads (Facebook & Instagram)
Default: 7-day click, 1-day view
Meta's default is shorter than Google's, which already creates a measurement mismatch. If the same person clicks a Google ad and a Facebook ad before converting, both platforms will claim the conversion — but with different attribution windows, they might not even agree on whether the conversion happened at all.
TikTok Ads
Default: 7-day click, 1-day view (same as Meta)
LinkedIn Ads
Default: 30-day click (no view-through tracking by default)
See the problem? If you're running campaigns across multiple platforms, you're comparing numbers that are fundamentally measuring different things.
Why Last-Click Attribution Breaks Down
Most platforms default to last-click attribution, which means the last ad someone clicked before converting gets 100% of the credit.
Here's a real customer journey:
- Monday: User sees your Facebook ad (doesn't click)
- Tuesday: User clicks your Google Search ad, visits site, leaves
- Wednesday: User sees a retargeting ad on Instagram (doesn't click)
- Thursday: User types your brand name into Google, clicks the ad, and buys
Under last-click attribution, that final Google branded search ad gets 100% of the credit. The Facebook ad that introduced your brand? Zero credit. The first Google ad that brought them to your site? Zero credit. The retargeting ad that reminded them you exist? Zero credit.
If you optimize your budget based on last-click data, you'll dump all your money into branded search (because it "converts the best") and kill the top-of-funnel ads that are actually generating the awareness that makes branded search work in the first place.
The Right Attribution Window for Your Business
There's no universal right answer, but there are smart guidelines based on your sales cycle.
Ecommerce (Impulse Products)
If you sell products people buy on impulse — apparel, accessories, consumables — most conversions happen within hours to a couple of days of the first click.
Recommended: 3-day click, 1-day view
This captures the customers who saw your ad, thought about it, came back the next day and bought. It doesn't give you credit for conversions that happened a week later and were probably driven by something else.
Considered Purchases (High-Ticket Items)
If you sell furniture, electronics, or anything over $500 that people research before buying, the cycle is longer.
Recommended: 7-day click, 1-day view
This gives people time to compare options, read reviews, and come back when they're ready to buy. Beyond 7 days, attribution gets fuzzy — did your ad really drive that purchase, or did they just happen to click your ad weeks ago and then buy for unrelated reasons?
B2B / Lead Generation
If you're generating leads for high-value B2B services, the journey from ad click to demo request can span days or even weeks.
Recommended: 14-day click, 1-day view for the lead event, then track offline conversions (closed deals) separately
The key here is to NOT use a long attribution window for the closed deal itself. Instead, track the initial lead with a 14-day window, then use offline conversion imports or CRM integrations to see which leads turned into revenue. That gives you a much clearer picture than trying to attribute a deal that closed 60 days after an ad click.
View-Through Attribution: Handle With Care
View-through attribution gives credit to ads people saw but didn't click, as long as they converted within the view window (usually 1 day).
The theory: someone saw your ad, didn't click, but the ad planted a seed. Later that day they came back via another channel and converted. The ad deserves some credit.
The reality: view-through attribution is noisy. Did the user actually see the ad, or did it load in a background tab they never looked at? Did the ad influence their decision, or would they have converted anyway?
Most sophisticated advertisers either turn off view-through entirely or set it to 1 day and treat it as a secondary signal, not a primary performance metric.
When to use view-through
- High-frequency awareness campaigns where you know users are seeing your ads repeatedly
- Retargeting campaigns where you're reminding people about products they already viewed
- Video campaigns where view-through makes more sense because video ads are harder to ignore than banner ads
When to ignore view-through
- Performance campaigns where you're optimizing for direct response
- Any campaign where you're making budget decisions based on ROAS — view-through inflates your reported conversions and makes ROAS look better than it is
Data-Driven Attribution: The Smart Alternative to Last-Click
Google Ads offers data-driven attribution (DDA), which uses machine learning to assign fractional credit to each touchpoint in a customer journey based on how much each one actually contributed to the conversion.
Instead of giving the last click 100% of the credit, DDA might give:
- 30% to the initial discovery ad
- 20% to a retargeting ad
- 50% to the final branded search click
The algorithm figures this out by comparing thousands of converting and non-converting paths and identifying which touchpoints statistically increase the likelihood of conversion.
If your account has at least 300 conversions in the past 30 days, you should be using data-driven attribution instead of last-click. It's not perfect, but it's dramatically better than pretending the last click is the only thing that mattered.
How to switch to data-driven attribution in Google Ads
- Go to Tools → Conversions
- Click the conversion action you want to change
- Scroll to Attribution model
- Select Data-driven
- Save
Your historical data will be reprocessed with the new model, so you'll see your conversion numbers shift (usually slightly downward for branded search, slightly upward for top-of-funnel campaigns).
Cross-Platform Attribution: The Unsolvable Problem
Here's the uncomfortable truth: if you're running ads on Google, Meta, TikTok, and LinkedIn simultaneously, there is no way to get a unified, accurate view of which platform deserves credit for which conversions.
Each platform only sees its own touchpoints. Google has no idea someone saw a TikTok ad before clicking your Google ad. TikTok has no idea they later converted via a Facebook retargeting ad. Each walled garden reports conversions based on its own attribution window and its own last-click (or data-driven) model.
The result: your conversions will add up to more than 100% of reality. If you had 100 actual sales, Google might claim 70, Meta might claim 60, and TikTok might claim 20 — because the same customer journey touched multiple platforms, and each one is taking credit.
How to handle it
You have two practical options:
Option 1: Use a single source of truth
Pick one platform (usually Google Analytics or your ecommerce backend) as the canonical source for conversion counts and ROAS. Use in-platform reporting only for optimization signals (which ads are performing within each platform), not for cross-platform budget allocation.
Option 2: Incremental testing
Run holdout tests where you turn off one platform for a defined period and measure the impact on overall revenue. This tells you how much incremental value each platform is actually delivering, independent of attribution models.
For example: turn off Meta ads for two weeks. If your total revenue drops by $10,000, Meta was contributing at least $10,000. If revenue drops by $2,000, Meta's in-platform reporting was overcounting its impact.
What About Post-Purchase Surveys?
One of the most underrated attribution methods: just ask your customers.
Add a simple post-purchase survey question: "How did you first hear about us?"
Options:
- Google search
- Facebook/Instagram ad
- TikTok
- Friend or family
- Other
Survey data isn't perfect (people forget, misremember, or don't pay attention to where they saw an ad), but it gives you a qualitative signal that complements your quantitative attribution data.
If 40% of survey respondents say they first heard about you from TikTok, but TikTok is only getting credit for 10% of conversions in your attribution reports, that's a strong signal you're under-investing in TikTok.
Common Attribution Mistakes
A few mistakes account for most broken attribution setups:
Using different windows across platforms. If Google is set to 30-day click and Meta is set to 7-day click, you can't compare their performance. Standardize your windows.
Ignoring view-through inflation. If you're optimizing for ROAS and including view-through conversions, your ROAS is artificially inflated. Either exclude view-through or report it separately.
Not using data-driven attribution when you qualify. If you have the conversion volume, last-click is just leaving money on the table.
Treating attribution as truth. Attribution models are models — they're useful approximations, not facts. Use them to guide decisions, but validate with incrementality tests and real revenue data.
The Practical Attribution Setup for Most Businesses
If you're not sure where to start, here's a sensible default configuration:
Shorten your attribution windows to match your actual sales cycle. Most businesses should be using 3-7 day click windows, not 30.
Enable data-driven attribution in Google Ads if you have the volume.
Disable or heavily discount view-through unless you're running high-frequency video campaigns.
Use GA4 or your backend as your source of truth for total conversions and revenue, not in-platform numbers.
Run quarterly incrementality tests to validate that your attribution models align with reality.
Wrapping Up
Attribution is hard because customer journeys are messy. People don't click one ad and immediately buy. They see your brand multiple times across multiple channels, they research, they forget, they come back, they compare, and then maybe they convert.
No attribution model captures that perfectly. Last-click over-credits bottom-of-funnel. Long attribution windows over-credit everything. Data-driven attribution is better but still imperfect. Cross-platform attribution is fundamentally unsolvable with today's tools.
The goal isn't perfect attribution — it's good enough attribution that you can make better budget decisions than you would with no data at all. Shorten your windows, switch to data-driven models where you can, discount view-through, and validate everything against real revenue. That's how you stop optimizing toward fiction and start optimizing toward profit.
